Load Data

data <- read.table("HPRD_Release9_062910/BINARY_PROTEIN_PROTEIN_INTERACTIONS.txt", 
                   header = F, 
                   sep ="\t",
                   col.names=c("GS1","HID1","Ref1","GS2","HID2","Ref2", "ExpT", "PID"))
#NOTE: The column names are as specified in README
#1. Interactor 1 Gene symbol
#2. Interactor 1 HPRD id
#3. Interactor 1 RefSeq id
#4. Interactor 2 Gene symbol
#5. Interactor 2 HPRD id
#6. Interactor 2 RefSeq id
#7. Experiment type (in vivo, in vitro and yeast 2-hybrid)
#8. Pubmed id

Data Processing

# Sort By Interactior 1 HPRD id followed by Interator 2 HPRD id
# Some gene symbols are missing so we rely on HPRD id
data <- data[order(data[,2], data[,5]),]
# Total nodes = Union of Unique IDs found in Interactor 1 or 2
totalNodes <- length(union(unique(data$HID1), unique(data$HID2)))
# Total edges = Number of rows in dataset(each row represents one edge, but not necessarily two nodes)
totalEdges <- nrow(data)
cat( paste("Total Nodes ", totalNodes, collapse=":") )

Total Nodes 9673

cat( paste("Total Edges ", totalEdges, collapse=":") )

Total Edges 39240

Calculate Out and In Degrees

transformed <- transform(data, OutDegree = ave(seq(nrow(data)), HID1, FUN=length))
transformed <- transform(transformed, InDegree = ave(seq(nrow(data)), HID2, FUN=length))

outDegrees <- transformed$OutDegree[!duplicated(transformed$HID1)]
inDegrees <- transformed$InDegree[!duplicated(transformed$HID2)]
cat( paste("Average out degree ", mean(outDegrees), collapse = ":") )

Average out degree 4.92346298619824

cat( paste("Average in degree ", mean(inDegrees), collapse = ":") )

Average in degree 5.21115537848606

cat( paste("Max in degree ", max(inDegrees), collapse = ":") )

Max in degree 175

cat( paste("Max out degree ", max(outDegrees), collapse = ":") )

Max out degree 173

cat( paste("Min in degree ", min(inDegrees), collapse = ":") )

Min in degree 1

cat( paste("Min out degree ", min(outDegrees), collapse = ":") )

Min out degree 1

Distribution of Out Degrees

od <- as.data.frame(table(outDegrees))
od$x<-as.numeric(od$outDegrees)
od$y<-od$Freq
#print(od, row.names=FALSE)
library(ggplot2)
ggplot(od, aes(x=as.numeric(od$outDegrees), y=od$Freq/sum(od$Freq))) + geom_point(shape=20) +xlab("log10(Degree)") + ylab("log10(Pr(Degree)") + scale_y_log10() + scale_x_log10() + geom_text(aes(x = 20, y = 50, label = lm_eqn(od)), parse = TRUE)

Distribution of In Degrees

id <- as.data.frame(table(inDegrees))
id$x<-as.numeric(id$inDegrees)
id$y<-id$Freq
#print(id, row.names=FALSE)
ggplot(id, aes(x=as.numeric(id$inDegrees), y=id$Freq/sum(id$Freq))) + geom_point(shape=20) +xlab("log10(Degree)") + ylab("log10(Pr(Degree)") + scale_y_log10() + scale_x_log10() + geom_text(aes(x = 20, y = 50, label = lm_eqn(id)), parse = TRUE)